Can Anthropic Keep Its Exploit-Writing AI Out of the Wrong Hands?
Its Mythos Preview model, which can allegedly find and exploit critical zero-days, also comes with certain controls, the vendor said.
Related Vulnerabilities
- CVE-2026-34971: Wasmtime: Miscompiled guest heap access enables sandbox escape on aarch64 Cranelift MEDIUM
- CVE-2026-28704: Emocheck insecurely loads Dynamic Link Libraries (DLLs). If a crafted DLL file is placed to the same HIGH
- CVE-2026-35621: OpenClaw before 2026.3.24 contains a privilege escalation vulnerability where the /allowlist command HIGH
- CVE-2026-34481: Apache Log4j's JsonTemplateLayout https://logging.apache.org/log4j/2.x/manual/json-template-layout. MEDIUM
- CVE-2026-35601: Vikunja has iCalendar Property Injection via CRLF in CalDAV Task Output MEDIUM
Related Coverage
Threat Actors